The peroxidase (POD)-like nanozymes are regarded as facile and convenient toolkits for the quantification of ascorbic acid (AA). Nevertheless, their insufficient specificity for AA in comparison to other biological antioxidants, such as glutathione (GSH) and cysteine (Cys), severely hindered the discrimination of AA in complex systems. To attain simultaneous quantification and discrimination of AA from GSH and Cys, a CuS nanoparticles-based sensor array with multichannel output signals is constructed in this study. The CuS nanoparticles exhibit superior POD-like activity and can effectively initiate the oxidation of 3, 3′, 5, 5′-tertamethylbenzidine (TMB) to oxidized TMB (oxTMB), accompanied by a measurable change in UV–Vis absorption, solution color and photothermal performance. However, in the presence of antioxidants, the aforementioned process is inhibited and the oxTMB is reduced back to TMB, resulting in signal (UV–Vis absorption, RGB and photothermal heating temperature) changes corresponding to the identity and concentration of antioxidants. As a consequence, by creating a finger print pattern using the multichannel output signals through linear discriminant analysis (LDA) and principal component analysis (PCA), the CuS nanoparticle-based sensor array enables the simultaneous quantification and discrimination of AA from GSH and Cys both in aqueous solution and human serum. This study presents a novel approach for the precise quantification and differentiation of AA and shows great potential in practical application.